Advanced Textbooks in Control and Signal Processing Series Editors Professor Michael J. Grimble, Professor of Industrial Systems and Director Professor Michael A. Johnson, Professor Emeritus of Control Systems and Deputy Director Industrial Control Centre, Department of Electronic and Electrical Engineering, University of Strathclyde, Graham Hills Building, 50 George Street, Glasgow G1 1QE, UK Other titles published in this series: Genetic Algorithms K.F. Man, K.S. Tang and S. Kwong Neural Networks for Modelling and Control of Dynamic Systems M. Nørgaard, O. Ravn, L.K. Hansen and N.K. Poulsen Modelling and Control of Robot Manipula to rs (2nd Edition) L. Sciavicco and B. Siciliano Faul t Detection and Diagnosis in Industrial Systems L.H. Chiang, E.L. Russell and R.D. Braatz Soft Computing L. Fortuna, G. Rizzotto, M. Lavorgna, G. Nunnari, M.G. Xibilia and R. Caponetto Statistical Signal Processing T. Chonavel Discrete-time Stochastic Processes (2nd Edition) T. Söderström Par a llel Computing for R eal-time Signal Processing and Control M.O. Tokhi, M.A. Hossain and M.H. Shaheed Multivariable Control Systems P. Albertos and A. Sala Control Systems with Input and Output Constraints A.H. Glattfelder and W. Schaufelberger Analysis and C ontrol of Non-linear Process Systems K. Hangos, J. Bokor and G. Szederkényi Model Predictive Control (2nd Edition) E.F. Camacho and C. Bordons Principles o f Adaptive Filters and Self-learning Systems A. Zaknich Digital Self-tuning Controllers V.Bobál,J.Böhm,J.FesslandJ.Machá ˇ cek Robust Control Design with MATLAB® D W. Gu, P.Hr. Petkov and M.M. Konstantinov Publication due July 2005 Active Noise and Vibration Control M.O. Tokhi Publication due November 2005 R. Kelly, V. Santibáñez and A. Loría Control of Robot Manipulators in Joint Space With 110 Figures 123 Rafael Kelly, PhD Centro de Investigación Científica y de Educación Superior de Ensenada (CICESE), Ensenada B.C. 22800, M exico Victor Santibáñez Davila, PhD Instituto Tecnologico de la Laguna, Torreón, Coahuila, 27001, Mexico Antonio Loría, PhD CNRS, Laboratoire des Signaux et Systèmes, Supélec, 3 rue J oliot Curie, 91192 Gif-sur-Yvette, France British Library Cataloguing in Publication Data Kelly, R. Control of robot manipulators in joint space. - (Advanced textbooks in control and signal processing) 1. Robots - Control systems 2. Manipulators (Mechanism) 3. Programmable con trollers I. Title II. Santibáñez, V. III. Loría, A. 629.8’933 ISBN-10: 1852339942 Library of Congress Control Number: 2005924306 Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or b y a ny means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. Advanced Textbooks in Control and Signal Processing series ISSN 1439-2232 ISBN-10: 1-85233-994-2 ISBN-13: 978-1-85233-994-4 Springer Science+Business M edia springeronline.com © Springer-Verlag London Limited 2005 The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore free for general use. The publisher makes no representation, express or implied, with regard to the accuracy of the infor- mation contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. Typesetting: Camera ready by authors Production: LE-T E XJelonek,Schmidt&VöcklerGbR,Leipzig,Germany Printed in Germany 69/3141-543210 Printed on acid-free paper SPIN 11321323 To my parents, with everlasting love, respect and admiration. –AL “Attentive readers, who spread their thoughts among themselves, always go beyond the author” —Voltaire ∗ , 1763. ∗ Original citation in French: “Des lecteurs attentifs, qui se communiquent leurs pens´ees, vont toujours plus loin que l’auteur”,inTrait´e sur la tol´erence `a l’occasion de la mort de Jean Calas, Voltaire, 1763. Series Editors’ Foreword The topics of control engineering and signal processing continue to flourish and develop. In common with general scientific investigation, new ideas, concepts and interpretations emerge quite spontaneously and these are then discussed, used, discarded or subsumed into the prevailing subject paradigm. Sometimes these innovative concepts coalesce into a new sub-discipline within the broad subject tapestry of control and signal processing. This preliminary battle be- tween old and new usually takes place at conferences, through the Internet and in the journals of the discipline. After a little more maturity has been acquired by the new concepts then archival publication as a scientific or engineering monograph may occur. A new concept in control and signal processing is known to have arrived when sufficient material has evolved for the topic to be taught as a specialized tutorial workshop or as a course to undergraduate, graduate or industrial engineers. Advanced Textbooks in Control and Signal Processing are designed as a vehicle for the systematic presentation of course material for both popular and innovative topics in the discipline. It is hoped that prospective authors will welcome the opportunity to publish a structured and systematic presentation of some of the newer emerging control and signal processing technologies in the textbook series. One of our aims for the Advanced Textbooks in Control and Signal Pro- cessing series is to create a set of course textbooks that are comprehensive in their coverage. Even though a primary aim of the series is to service the textbook needs of various types of advanced courses we also hope that the industrial control engineer and the control academic will be able to collect the series volumes and use them as a reference library in control and signal processing. Robotics is an area where the series has the excellent entry in the volume by L. Sciavicco and B. Siciliano entitled Modelling and Control of Robot Ma- nipulators, now in its second edition. To complement our coverage in Robotics, we are pleased to welcome into the series this new volume Control of Robot Manipulators in Joint Space by Rafael Kelly, V´ıctor Santib´a˜nez and Antonio Lor´ıa. Other topics like models, kinematics and dynamics are introduced into x Series Editors’ Foreword the narrative as and when they are needed to design and compute the robot manipulator controllers. Another novel feature of the text is the extensive use of the laboratory prototype P elican robotic manipulator as the test-bed case study for the robot manipulator controllers devised. This ensures that the reader will be able to see how robot manipulator control is done in practice. Indeed, this means that the text can be closely linked to “hands on” laboratory experience. Control and mechatronics lecturers wishing to use the textbook to support their advance course on robot manipulator control will find the lecture presentation slides, and the problem solutions, which are available at springonline.com, an added bonus. The style of the text is formally rigorous but avoids a lemma–theorem presentation in favour of one of thorough explanation. Chapter 2 of the text covers the main mathematical tools and introduces the concepts of the direct (or second) method of Lyapunov for system stability analysis. This is needed because the robot manipulator system is a nonlinear system. Since the cover- age in this chapter includes a wide range of stability concepts, the reader will be pleased to find each new concept supported by a worked example. Robot dynamics and their implications for robot manipulator control are covered in Chapters 3 and 4 whilst Chapter 5 moves on to discuss the model details of the Pelican prototype robotic manipulator. The kinematic and dynamic mod- els are, described and model parameter values given. This chapter shows how the Pelican prototype is “kitted out” with a set of models the properties of which are then investigated in preparation for the control studies to follow. Parts II to IV (covering Chapters 6 to 16) are devoted to robot manip- ulator controller design and performance case studies. This shows just how focused the textbook is on robot manipulator control. This study is given in three stages: position control (Part II); motion control (Part Ill) and ad- vanced control topics (Part IV). Remarkably, the workhorse controller type being used is from the PID family so that the control focus is close to the type of controllers used widely in industrial applications, namely from the classical Proportional, Integral, Derivative controller family. In these chapter- length controller studies, the earlier lessons in Lyapunov stability methods come to the fore, demonstrating how Lyapunov theory is used for controllers of a classical form being used with nonlinear system models to prove the necessary stability results. The advanced control topics covered in Part IV include a range of adaptive control methods. Four appendices are given with additional material on the mathematical and Lyapunov methods used and on the modelling details of direct current motors. There is no doubt that this robot manipulator control course textbook is a challenging one but ultimately a very rewarding one. From a general viewpoint the reward of learning about how to approach classical control for systems having nonlinear models is a valuable one with potential application in other control fields. For robot manipulator control per se, the book is rigorous, thorough and comprehensive in its presentation and is an excellent addition to the series of advanced course textbooks in control and signal processing. M.J. Grimble and M.A. Johnson Glasgow, Scotland, U.K. March 2005 Preface The concept of robot has transformed from the idea of an artificial super- human, materialized by the pen of science fiction writer Karel ˇ Capek, into the reality of animated autonomous machines. An important class of these are the robot manipulators, designed to perform a wide variety of tasks in production lines of diverse industrial sectors; perhaps the most clear exam- ple is the automotive industry. Robotics, introduced by science fiction writer Isaac Asimov as the study of robots, has become a truly vast field of modern technology involving specialized knowledge from a range of disciplines such as electrical engineering, mechatronics, cybernetics, computer science, mechani- cal engineering and applied mathematics. As a result, courses on robotics continue to gain interest and, following the demands of modern industry, every year more and more program studies, from engineering departments and faculties of universities round the globe, include robotics as a compulsory subject. While a complete course on robotics that is, including topics such as modeling, control, technological implementation and instrumentation, may need two terms at graduate level to be covered in fair generality, other more specialized courses can be studied in one senior year term. The present text addresses the subject in the second manner; it is mostly devoted to the specific but vast topic of robot control. Robot control is the spine of robotics. It consists in studying how to make a robot manipulator do what it is desired to do automatically; hence, it consists in designing robot controllers. Typically, these take the form of an equation or an algorithm which is realized via specialized computer programs. Then, controllers form part of the so-called robot control system which is physically constituted of a computer, a data acquisition unit, actuators (typically elec- trical motors), the robot itself and some extra “electronics”. Thus, the design and full implementation of a robot controller relies on every and each of the above-mentioned disciplines. The simplest controller for industrial robot manipulators is the Propor- tional Integral Derivative (PID) controller. In general, this type of controller xii Preface is designed on the basis that the robot model is composed of independent cou- pled dynamic (differential) equations. While these controllers are widely used in industrial manipulators (robotic arms), depending on the task to be carried out, they do not always result in the best performance. To improve the latter it is current practice to design so-called model-based controllers, which require a precise knowledge of the dynamic model including the values of the physi- cal parameters involved. Other, non-model-based controllers, used mainly in academic applications and research prototypes include the so-called variable- structure controllers, fuzzy controllers, learning controllers, neural-net-based controllers, to mention a few. The majority of available texts on robotics cover all of its main aspects, that is, modeling (of kinematics and dynamics), trajectory generation (that is, the mathematical setting of a task to be performed by the robot), robot control and some of them, instrumentation, software and other implementation issues. Because of their wide scope, texts typically broach the mentioned topics in a survey rather than a detailed manner. Control of robot manipulators in joint space is a counter-fact to most avail- able literature on robotics since it is mostly devoted to robot control, while ad- dressing other topics, such as kinematics, mainly through case studies. Hence, we have sacrificed generality for depth and clarity of exposition by choosing to address in great detail a range of model-based controllers such as: Pro- portional Derivative (PD), Proportional Integral Derivative (PID), Computed torque and some variants including adaptive versions. For purely didactic rea- sons, we have also chosen to focus on control in joint space, totally skipping task space and end-effector space based control. These topics are addressed in a number of texts elsewhere. The present book opens with an introductory chapter explaining, in gen- eral terms, what robot control involves. It contains a chapter on preliminaries which presents in a considerably detailed manner the main mathematical con- cepts and tools necessary to study robot control. In particular, this chapter introduces the student to advanced topics such as Lyapunov stability, the core of control theory and therefore, of robot control. We emphasize at this point that, while this topic is usually reserved for graduate students, we have paid special attention to include only the most basic theorems and we have reformulated the latter in simple statements. We have also included numer- ous examples and explanations to make this material accessible to senior year undergraduate students. Kinematics is addressed mainly through examples of different manipula- tors. Dynamics is presented in two chapters but from a viewpoint that stresses the most relevant issues for robot control; i.e. we emphasize certain funda- mental properties of the dynamic model of robots, which are commonly taken as satisfied hypotheses in control design. [...]... 10 7 q 4.4 Problem 1 11 0 5 .1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 II .1 II.2 Pelican: experimental robot arm at CICESE, Robotics lab 11 4 Diagram of the 2-DOF Pelican prototype robot 11 4 Two solutions to the inverse kinematics problem 11 7 “Bent-over” singular configuration 11 9 Desired reference... position errors against time 237 Problem 1 Cartesian 2-DOF robot 239 11 .1 11. 2 11 .3 11 .4 11 .5 Block-diagram: PD control with compensation 245 Diagram of the Pelican robot 247 Graph of position errors against time 248 Block-diagram: PD+ control 249 Graph of... 11 0 5 Case Study: The Pelican Prototype Robot 11 3 5 .1 Direct Kinematics 11 5 5.2 Inverse Kinematics 11 6 5.3 Dynamic Model 11 9 5.3 .1 Lagrangian Equations 11 9 5.3.2 Model in Compact Form 12 3 5.4 Desired... 309 14 Introduction to Adaptive Robot Control 313 14 .1 Parameterization of the Dynamic Model 314 Contents xxi 14 .1. 1 Linearity in the Dynamic Parameters 315 14 .1. 2 The Nominal Model 319 14 .2 The Adaptive Robot Control Problem 325 14 .3 Parameterization of the Adaptive... 205 Diagram of the Pelican prototype robot 214 ˜ Graphs of position errors q1 and q2 215 ˜ ˜ Graphs of position errors q1 and q2 216 ˜ 10 .1 10.2 10 .3 10 .4 10 .5 Block-diagram: computed-torque control 228 Graph of position errors against time 232 Computed-torque+ control ... 418 Index 419 List of Figures I .1 Robot manipulator 1. 1 1. 2 1. 3 1. 4 1. 5 1. 6 Freely moving robot 8 Robot interacting with its environment 8 Robotic system: fixed camera 9 Robotic system: camera... Diagram of a robot with elastic joints 78 3.8 3.9 Link with an elastic joint 81 Example of a 2-DOF robot 81 3 .10 Example of a 2-DOF robot 82 3 .11 Diagram of a DC motor 83 3 .12 Block-diagram of a robot with its actuators 84 3 .13 Pendular... 6.5 Block-diagram: Proportional control plus velocity feedback 14 1 Block-diagram: PD control 14 2 Graph of q (t)2 15 1 ˜ Graph of q(t)2 15 1 ˙ ˜ Graph of the position errors q1 and q2 15 2 ˜ 7 .1 Block-diagram: PD control with gravity compensation 15 8 List... 270 ˜ 12 .5 Graphs of position errors q1 and q2 282 ˜ 13 .1 Block-diagram: P“D” control with gravity compensation 293 13 .2 Graphs of position errors q1 (t) and q2 (t) 299 ˜ ˜ 13 .3 Block-diagram: P“D” control with desired gravity compensation 300 ˜ 13 .4 Graphs of position errors q1 (t) and q2 (t) 307 ˜ 14 .1 Planar 2-DOF manipulator... 322 14 .2 Block-diagram: generic adaptive control of robots 329 14 .3 Problem 2 Cartesian robot 335 ˜ 15 .1 Graphs of position errors q1 and q2 355 ˜ 1 and θ2 356 ˆ 15 .2 Graphs of adaptive parameters θ 15 .3 Graphs of position errors q1 and q2 357 ˜ ˜ ˆ ˆ 15 .4 Graphs of adaptive parameters θ1 . and Signal Processing series ISSN 14 3 9-2 232 ISBN -1 0 : 1- 8 523 3-9 9 4-2 ISBN -1 3 : 97 8 -1 -8 523 3-9 9 4-4 Springer Science+Business M edia springeronline.com © Springer-Verlag London Limited 2005 The use. 10 8 Bibliography . 10 9 Problems . 11 0 5 Case Study: The Pelican Prototype Robot 11 3 5 .1 Direct Kinematics 11 5 5.2 Inverse Kinematics 11 6 5.3 Dynamic Model . . 11 9 5.3 .1 Lagrangian Equations . 11 9 5.3.2. 417 Bibliography . 418 Index 419 List of Figures I .1 Robot manipulator 4 1. 1 Freely moving robot . . . 8 1. 2 Robot interacting with its environment. . 8 1. 3 Robotic system: fixed camera . . 9 1. 4